Habitat is an extremely complex natural complex,which is formed by the combination of many habitat factors,including climate,terrain,soil,biology and so on.Classifying habitats helps reveal the complexity of nature.By classifying habitats,they can be described and compared more effectively,and biological conservation or ecological restoration work can be carried out according to the special characteristics of the habitats.The loess hilly region has complex and varied topography and numerous habitat types.How to select suitable species for different habitats has always been one of the key problems to be solved in vegetation restoration in this region.Based on the classification of habitat types,this project established the spatial distribution of functional traits of plant communities to clarify the characteristics and thresholds of different habitats,providing strong support for habitat species screening.In this study,the spatial distribution of terrain factors and climate factors was obtained by extraction of terrain factors and spatial interpolation of climate factors.Then,the spatial distribution of soil factors was obtained by inversion according to the response relationship model between soil moisture and nutrient data and topographic and climatic factors,and the spatial distribution characteristics of habitat factors in the loess hilly region were revealed.Secondly,plant functional data were obtained through investigation and sampling,community weighted character values of plant functional traits were calculated,and regression analysis was conducted with environmental factors to explore the driving mechanism of plant community functional traits change,reveal the multi-scale effects of environmental factors on plant community functional traits,and obtain the spatial distribution of seven plant community functional traits.Finally,the K-means clustering algorithm was introduced to classify the habitat types in the loess hilly region,and the distribution characteristics of plant functional traits in different habitats were discussed.The main conclusions reached are as follows:(1)The variation characteristics of the main habitat factors in the loess hilly region were analyzed.According to the spatial distribution of the topographic factors,the topographic factors were extracted.The surface roughness range was 1-4.19Ra,the topographic relief range was 0-233m,and the slope range was 0-66.135°.According to ANUSPLIN interpolation,the average annual temperature ranges from 7.725℃to 13.329℃,the average annual extreme high temperature ranges from 14.929℃to 20.968℃,the average annual low temperature ranges from 1.737℃to 8.136℃,and the average daily annual precipitation ranges from 1.018mm to 2.042mm.For the spatial distribution of water nutrients in the loess hilly region,the soil moisture content was relatively low,with the maximum soil moisture content being10.87%,the minimum soil moisture content being 2.45%,and the average soil moisture content being 8.28%.The spatial inversion range of soil total nitrogen content is 0.25~1.77g·kg-1,the soil total phosphorus content is 0.43~0.60 g·kg-1,and the soil organic matter content is 2.49~26.96 g·kg-1.On the whole,the soil nutrient is medium low level.(2)The regional variation characteristics of plant functional traits in the Loess hilly region were identified:Except root volume,the other seven plant functional traits were significantly correlated with environmental factors(P<0.05),through the inversion of the linear regression model of plant functional traits and environmental factors,so that each plant functional traits can be spatialized.Leaf tissue density ranged from 0.06 to 1.44 g·cm-3,and leaf phosphorus ranged from 0.18 to 2.85 mg·g-1.The two factors showed roughly the same rule,from east to west,from small to large.Leaf nitrogen ranged from 7.11 to 25.02 mg·g-1,which was higher in the east,followed by the northwest,and lowest in the south.The leaf carbon content ranged from 410.19 to 520.79 mg·g-1,and the leaf carbon content was from high to low from east to west,showing obvious hierarchy.The density of root tissue ranged from 0.22 to 1.04 g·cm-3.The density of root tissue was higher in the northwest and eastern parts of the region,but smaller in the middle part.The leaf thickness ranges from 0.10 mm to 0.25 mm,with large values concentrated in the east and south,and small values in the northwest and a few central areas.The specific leaf area ranges from 54.18 to 213.72 cm2·g-1,with a large span.The specific leaf area is the largest in the east,the second in the middle,and the smallest in the northwest.(3)The K-means clustering algorithm was introduced to classify the habitat types in the loess hilly region,and the distribution range and threshold of functional traits of each habitat were defined.The region was divided into 5 habitats by K-means clustering based on climate factor,topographic factor and soil factor.By superposition analysis of habitat type map and plant functional traits distribution,the distribution range and threshold of functional traits in each habitat were defined.In Habitat 1,soil moisture,soil organic matter,soil total nitrogen,soil total phosphorus,and average annual daily precipitation were low,and average annual temperature,average annual extreme high temperature,and average annual extreme low temperature were all high.The environment was in a poor condition,and the root tissue density of vegetation community was 0.43~1.44 g·cm-3 and leaf thickness was 0.14~0.18 mm,which were high values.The root volume was 0.15~1.44 m L,the specific leaf area was 100~146cm2·g-1,the leaf tissue density was 0.05~0.84 g·cm-3,and the leaf phosphorus content was0.18~1.50 mg·g-1.The plant community adapted to the environment by changing these functional traits.However,the functional traits of habitat 3 and 4 with better living environment were opposite to those of habitat 1. |